Statistical Robustness, Stability and Interpretability of Algorithms

Interpretable machine learning and arti_cial intelligence is a big emerging theme, complementing the development of pure black box prediction tools.

Looking through the lens of statistical causality opens up new paths and opportunities for enhanced interpretability and predictive robustness of algorithms, with wide-ranging prospects for various applications; and we will highlight some of them from the area of molecular biology. The key technical idea relies on a notion of probabilistic invariance, exhibiting novel connections to robust optimization.

Technical Secretariat

logoGPBlanco Marià Cubí 4 - Pral
08006 Barcelona (SPAIN)
T. +34 932388777
F. +34 932387488
ICIAM2019 | All rights Reserved.